Performance Max Vs. Search: A Non-Scientific Look at If It Competes and What It Means
A brief look at Google’s Enhanced Campaigns backlash and the rise of Performance Max, highlighting lost control and advertiser concerns.
Performance Max Vs. Search: A Non-Scientific Look at If It Competes and What It Means
The 2013 Enhanced Campaigns fiasco was the greatest pushback on a Google rollout that effectively confiscated the ability to bid by device.
Then, Performance Max came.
Performance Max (also known as PMax) is considered to be a new ad unit introduced by Google, currently in beta testing with Bing, for those who are as blind to reality as a NY Jets fan.
Still being refused to call Microsoft Advertising. Many advertisers weren’t sure what to expect when it was launched in 2021.
Performance Max is a campaign types that enable advertisers to access all of Google’s ad inventory from one campaign. The inventory includes placements in – Search, Display, YouTube, Shopping, Discover, Maps (via local assets when applicable), and Gmail.
It works best for advertisers who are open to testing, have strong conversion signals, and are looking to scale beyond the limitations of a single channel.
Google does claim that PMax doesn’t cannibalize ad placements – which means if you run Search, Shopping, or another ad unit in parallel with PMax, they won’t compete.
Test Scenario
Prior to the brand wanting to pause PMax,
Consistently run brand search (Phrase and Extract match only) and PMax side by side.
Performance was often mixed and sometimes profitable, sometimes not. This results in a typical non-peak shopping season for their vertical.
Bid strategies were always constant:
- PMax on target return on ad spend (TROAS)
- one search campaign on TROAS
- one on max conversion value.
These campaigns consistently chugged along – although prior to the brand making the call, there was a billing issue with the brand’s credit card, so the account went down for three days (a short period of time that shouldn’t impact the algorithms).
PMax would get pulled offline for eight days, but for the sake of sanity purposes, seven days analysis of it offline vs. seven days before it went offline and seven days after it came back online.
Pre-PMax Offline Period
The three campaigns (two brand search and one PMax) had 185K impressions, with PMax driving 95% of them. There had been 3.2K clicks (not doing “Interactions” for this, for the sake of continuity), with search driving 72% of them.
Spend was a little under USD 4,900, and the aggregate cost per click (CPC) was USD 1.24.
Having 99 conversions, out of which 1/3 come from PMax
PMax Offline Period
Some of the interesting data was observed.
It should be noted that the search bid strategy was not changed, but the funds used for PMax were reallocated back to search.
The shift in setup had an impact, but not quite in the way you’d expect.
Search impressions rose 67%, but the aggregate impressions dropped 92%, as PMax wasn’t dumping the impressions anywhere.
Search clicks rose 92%, but aggregate rose just 39% (remember this 39% growth, because here is where it gets a bit mind-boggling).
Unrestricted by PMax stepping, search spending increased by an astounding 308%, but the overall increase was only 119%.
Although there was only a 39% increase in clicks, there was a significant increase in spend from the reallocation (from PMax to branded search), enabling search to attain its maximum spend potential (within budget constraints).
What caused this expenditure to soar? Well, the search CPCs grew 60%+ when PMax exited, driving the aggregate CPC up 51%.
Additionally, impression share lost to budget decreased by 615, impression share lost to rank decreased by 24%, and branded search impression share increased by 26%.
Relaunch of PMax
Reactivated PMax on the 9th day from when it was paused.
Its budget was sent back, which has been reallocated to search, but no other changes were made.
On week two of its relaunch, performance was evaluated and compared to the PMax offline period.
In branded search, an 11% drop in brand search impressions, a 23% drop in clicks, and a 24% drop in spend were observed, but search CPC remained flat.
Courtesy of PMax returning, aggregate impressions rose 35%, clicks 29%, and CPC dropped 7%. In addition, we saw a 91% increase in conversions.
What Does This Mean?
PMax has a clear and definitive impact on brand search when it is off.
With spending wavering, despite even reallocation of the aggregate budget back into brand search and then back into PMax, the data gets fuzzy.
- Brand search CPC is at its highest cost when it is solo (vs that of brand search and PMax)
- Brand search CTR is higher when PMax is offline
- Brand search CVR is lower when PMax is on
- There is an exchange of click volume between PMax and brand search when PMax is on vs. off.
Key Takeaways
Google defining “does not compete with paid search” is a bit suspect when they are run together.
As the loss/gain volume of clicks is a dead giveaway, there is an immediate impact.
At the same time, running them dually shows an aggregate net impact, positively for CVR and CPC.
Providing the positive impact of CPC is for brand search, and PMax’s CPC is equal to or lower than the share of clicks it takes from brand search, then who cares?
